Despite substantial research on adolescent tobacco use, we are far from having reliable, effective methods of preventing youth tobacco use onset, and prevalence of adolescent tobacco use in the U.S. remains high. The proposed research would test the mobilization of the powerfu1 influences of parents and peers by experimentally evaluating the efficacy of two strategies for preventing the onset of tobacco use in early adolescence. We propose to expand and strengthen the Family Communications About Tobacco and Youth Anti-tobacco Activities; interventions that were shown to have promising results in the previous project, to augment those interventions with videotapes and Internet technology, and to evaluate their combined effects when delivered to sixth graders. Over the first two years of the study, a total of 40 Oregon middle schools will be recruited, assessed on eight-grade prevalence of tobacco use, and randomly assigned to receive or not to receive the Family Communications and Youth Anti-tobacco interventions. Twenty schools will be recruited in the first year of the study, and 20 in the second year. The intervention targets sixth grade students; prevalence of tobacco use, other substance use, and antisocial behavior will be assessed among target students at the outset of sixth grade, and followed up at the end of those students' seventh and eighth grades. The study is also designed to assess exposure to intervention elements and the impact of the intervention on parenting practices and peer influence process the effect of the intervention on these variables will be measured via telephone interviews to a random sample of 50 child-parent pairs from each school, and relationships of parent and peer influences to subsequent tobacco use will be modeled using Latent Growth Modeling. Study design is a group-randomized design in which students are nested within schools, and intervention effects will be examined in two ways: using data from all students in grade 6 during intervention who stay in the school at subsequent assessments; and using data from students who transfer into the school in later grades.